We propose to measure the frequency dependence of respiratory impedance by forced random noise in dogs and man, and the transmural pressure-area characteristics of the larger airways in dogs. Normal, physiologically altered, and abnormal subjects, will be studied. Data will be processed using regression analysis and an algorithm based on the frequency dependence of effective resistance to compute values for anatomically meaningful parameters in respiratory models containing elements for airway compliance, central and peripheral resistance, and parallel parenchymal resistances and compliances. Parameter values from forced random noise will be correlated and compared with more standard parameters. The ultimate goal is to develop a rapid, simple-to-use, sensitive, non-invasive method for detecting and characterizing small changes in pulmonary mechanical functions during the early asymptomatic state of obstructive lung disease and to explore the role of airways compressibility in causing expiratoy flow limitation.